Dramatic Differences between the Energy Landscapes of SiO2 and

Sep 27, 2007 - Davy Faraday Research Laboratory, Royal Institution of Great Britain, 21 Albemarle Street, London W1S 4BS, U.K., and Department of Chem...
0 downloads 9 Views 79KB Size
Published on Web 09/27/2007

Dramatic Differences between the Energy Landscapes of SiO2 and SiS2 Zeotype Materials Martijn A. Zwijnenburg,*,† Furio Cora` ,†,‡ and Robert G. Bell†,‡ DaVy Faraday Research Laboratory, Royal Institution of Great Britain, 21 Albemarle Street, London W1S 4BS, U.K., and Department of Chemistry, UniVersity College London, 20 Gordon Street, London WC1H 0AJ, U.K. Received April 20, 2007; E-mail: [email protected]

Crystalline nanoporous materials known as zeotypes, including aluminosilicate zeolites, silicas, and aluminophosphates are of great industrial importance in a variety of technologies, such as catalysis, separation, and ion-exchange. The potentially enhanced range of applications for new structures with, for instance, larger pore sizes or novel electronic properties continues to drive the quest for new materials. The framework structures of zeotype materials are all based on 4-connected nets, and recent years have seen an enormous effort to enumerate, via theoretical approaches, all such nets.1 The 10000+ hypothetical nets so far generated show that the zeotype structural landscape is much richer than expected from the 176 currently synthesized or naturally occurring frameworks. However, it is predicted that the majority of these hypothetical frameworks are thermodynamically unfeasible as siliceous materials because they can only be realized with severely distorted TX4 tetrahedra2 and have heats of formation tens of kJ/mol (SiO2) higher than any currently hydrothermally synthesized material.1d,f Changes of the chemical composition to GeO2 or AlPO4 do not appear to alter this conclusion.2,3 Recently, however, the previously hypothetical RWY framework (where the T-atoms of the sodalite framework are replaced with supertetrahedral T4X6 units, see Figure 1), which we would deem unfeasible in a siliceous form, was synthesized as the gallium germanium sulfide material UCR-20.4 This inspired us to study in detail the energy landscape of tetrahedral sulfide (TS4) materials and compare it with that of silica. Here as a first step we carry out periodic density functional theory (DFT) calculations on SiS2 materials. The polymorphs studied included two known phases: tetragonal5 (isostructural to cristobalite, dia topology) and orthorhombic6 (consisting of a packing of tworing chains) and four model crystalline materials whose topologies correspond to the dense framework quartz (qtz) and the nanoporous frameworks SOD, CHA, and RWY. For comparison we also performed similar calculations on SiO2 versions of the same crystals. The total energies and optimized geometries of the different SiO2 and SiS2 materials were obtained using the PBE7 and PBE08 functionals and localized orbitals as implemented in the Crystal 06 code.9 Orthorhombic SiS2 was modeled as an isolated two-ring chain to circumvent problems with the poor description in DFT of the weak intrachain interactions. Frequencies of the γ-point phonons were calculated for each framework to verify that all optimized structures were true minima with respect to the internal coordinates. The latter is important since, with the exception of tetragonal and orthorhombic SiS2, the starting structures for the SiS2 materials could only be approximate and thus may optimize to n-order saddle points if left unchecked. Finally, the degree of distortion of the tetrahedra compared to an ideal tetrahedron (all X-T-X angles equal to 109.45° and all T-X bonds the same length) was assessed by the GASP code of Wells and co-workers.10 † ‡

Royal Institution of Great Britain. University College London.

12588

9

J. AM. CHEM. SOC. 2007, 129, 12588-12589

Figure 1. Fragment of RWY structure showing supertetrahedral units. Table 1. Total Energies Relative to Quartz, Total Tetrahedral Distortion, and Average Si-X-Si Angle for the PBE0 Optimized SiO2 and SiS2 Materials E (kJ/mol SiX2)

qtz dia SOD CHA RWY two-chain

〈Si−X−Si〉 (deg)

TTD (Å)

SiO2

SiS2

SiO2

SiS2

SiO2

SiS2

0 2 7 9 46

0 -33 -31 1 -37 -41

2.2 × 10-2 3.4 × 10-2 2.3 × 10-2 3.0 × 10-2 7.1 × 10-2

1.9 × 10-1 1.7 × 10-1 1.1 × 10-1 3.1 × 10-1 1.5 × 10-1 2.0 × 10-1

141 141 144 147 123

119 112 111 116 105 81

Table 1 gives the total energies (in kJ/mol relative to the respective quartz structure), total tetrahedral distortion values and average Si-X-Si angles of the different SiO2 and SiS2 materials, optimized using the PBE0 functional. The PBE functional gave essentially the same data, reinforcing our confidence in the results. While the silica materials show the expected energetic ordering with quartz as lowest energy structure, in line with experiment, the SiS2 energetic landscape appears to be dramatically different. So much so that, as sulfide, supertetrahedral RWY lies between the experimentally prepared tetragonal and orthorhombic SiS2 structures in terms of its thermodynamic stability. Furthermore, it is also much more stable than those frameworks which share the structure of low-energy silica polymorphs (e.g., quartz), strongly suggesting that sulfur holds the key to its successful synthesis. The other structure considered with small rings, the two-ring chain of orthorhombic SiS2, is also considerably stabilized when going from SiO2 to SiS2. The SiS2 two-ring chain is found to lie lowest in energy of all SiS2 materials consistent with the fact that it can be experimentally purified through sublimation6 and will be further stabilized by intrachain interactions not taken in account here, while its silica version (Silica-W11) was previously predicted to lie ∼120 kJ/mol SiO2 above quartz.12 The tetrahedral distortion values are given in Table 1 in terms of the rms length of the vectors linking the X atom positions of 10.1021/ja0727666 CCC: $37.00 © 2007 American Chemical Society

COMMUNICATIONS

Figure 2. Ratio of SiS2 and SiO2 pseudofrequencies calculated for quartz, plotted against the pseudofrequencies for the SiO2 structure.

Figure 3. Unit cell of dt2_10 hypothetical zeolite structure.

the tetrahedra in the optimized structures and their idealization. These demonstrate that the sulfide materials are all considerably more distorted than their silica analogues. In SiO2 form, values of the order of 2 × 10-2 Å are typical of experimentally known silica polymorphs such as quartz, cristobalite, or high-silica zeolites such as MFI.2 The RWY and two-chain structures have significantly higher values, outside the range observed in real siliceous materials and consistent with their much higher energies. Among the sulfide materials, a different picture emerges. The values of the tetrahedral distortion, including that of the experimentally known tetragonal SiS2 material, are between 1 × 10-1 and 3.5 × 10-1 Å, one order of magnitude higher than for SiO2. This implies that “normal” distortion in SiS2 is greater than in SiO2. Quartz, tetragonal SiS2, SOD, and CHA all have similar values, comparable to that of RWY, whereas as silica their relative distortion is much lower. It is also striking that RWY is even more distorted as sulfide than in a silica form, though without an analogous energy penalty. This increase in tetrahedral distortion is accompanied by an overall decrease of the intertetrahedral Si-XSi angles. The latter appears to reflect a preference of SiS2 for smaller Si-X-Si angles than silica. This analysis is supported by the fact that the data in Table 1 show an approximate correlation between the average Si-S-Si angle and the relative stability: the smaller the average Si-S-Si angle is (i.e., the richer the structure is in two and three rings), the more stable a SiS2 material is predicted to be. The dramatic change in the energy landscape is thus most likely due to the combination of a preference of sulfides for smaller intertetrahedral Si-X-Si angles and an increased tolerance of SiS2

for (tetrahedral) distortion compared to silica. The latter contribution is probably critical and the result of the smaller electrostatic cost due to the less ionic bonding in sulfides compared with oxides (as discussed by us in ref 13). The change in the relative ease with which the framework tetrahedra may be distorted is reflected in an increased flexibility of the material. This is seen clearly when comparing the pseudofrequencies of SiO2 and SiS2 quartz (i.e., the frequencies obtained with all the atomic masses set equal to one for comparison). The calculated pseudofrequencies are, depending on the particular mode in question, 10 to 50% lower for SiS2 than for SiO2 quartz (see Figure 2), suggesting a similar decrease in all generalized force constants. Our calculations thus strongly suggest that SiS2 and related sulfide compositions are ideally suited to the synthesis of frameworks whose structural units are inherently strained as SiO2, such as supertetrahedra, two- and three-rings; an example is the hypothetical framework dt2_101g (see Figure 3), which our calculations predicts to lie as sulfide at least 26 kJ/mol below quartz compared with 46 kJ/mol above quartz as silica. Targeting sulfide chemistry could therefore be highly productive in terms of synthesis of novel nanoporous structures, including the growing number of hypothetical nets generated computationally. Acknowledgment. We thank Drs. S.T. Bromley, R.A. Jackson, A.A. Sokol, and Prof. Dr. J. Sauer for stimulating discussion. M.A.Z. acknowledges the European Commission for a Marie Curie Intra-European Fellowship (MEIF-CT-2005-010326). Supporting Information Available: Full details of the DFT calculations. This material is available free of charge via the Internet at http://pubs.acs.org. References (1) (a) Treacy, M. M. J.; Randall, K. H.; Rao, S.; Perry, J. A.; Chadi, D. J. Z. Kristallogr. 1997, 212, 768. (b) Delgado Friedrichs, O.; Dress, A. W. M.; Huson, D. H.; Klinowski, J.; Mackay, A. L. Nature 1999, 400, 644. (c) Mellot-Draznieks, C.; Girard, S.; Ferey, G. J. Am. Chem. Soc. 2002, 124, 15326. (d) Foster, M. D.; Simperler, A.; Bell, R. G.; Delgado Friedrichs, O.; Almeida Paz, F. A.; Klinowski, J. Nat. Mater. 2004, 3, 234. (e) Treacy, M. M. J.; Rivin, I.; Balkovsky, E.; Randall, K. H.; Foster, M. D. Microporous Mesoporous Mater. 2004, 74, 121. (f) Foster, M. D.; Delgado Friedrichs, O.; Bell, R. G.; Almeida Paz, F. A.; Klinowski, J. J. Am. Chem. Soc. 2004, 126, 9769. (g) Simperler, A.; Foster, M. D.; Delgado Friedrichs, O.; Bell, R. G.; Almeida Paz, F. A.; Klinowski, J. Acta Crystallogr., Sect. B 2005, 61, 263. (h) Earl, D. J.; Deem, M. W. Ind. Eng. Chem. Res. 2006, 45, 5449. (2) Zwijnenburg, M. A.; Simperler, A.; Wells, S. A.; Bell, R. G. J. Phys. Chem. B 2005, 109, 14783. (3) Simperler, A.; Foster, M. D.; Bell, R. G.; Klinowski, J. J. Phys. Chem. B 2004, 108, 869. (4) Zheng, N.; Bu, X.; Wang, B.; Feng, P. Science 2002, 298, 2366. (5) Prewitt, C. T.; Young, H. S. Science 1965, 149, 535. (6) Peters, J.; Krebs, B. Acta. Crystallogr., Sect. B 1982, 38, 1270. (7) Perdew, J. P.; Burke, K.; Ernzerhof, M. Phys. ReV. Lett. 1996, 77, 3865. (8) Perdew, J. P.; Emzerhof, M.; Burke, K. J. Chem. Phys. 1996, 105, 9982. (9) Dovesi, R.; Saunders, V. R.; Roetti, C.; Orlando, R.; Zicovich-Wilson, C. M.; Pascale, F.; Civalleri, B.; Doll, K.; Harrison, N. M.; Bush, I. J.; D’Arco, Ph.; Llunell, M. CRYSTAL 06 User’s Manual, Universita’ di Torino: Torino, Italy, 2006. (10) Wells, S. A.; Dove, M. T.; Tucker, M. G. J. Appl. Crystallogr. 2004, 37, 536. (11) Weiss, A. Z. Anorg. Allg. Chem. 1954, 276, 95. (12) Bromley, S. T.; Moreira, I. D. P. R.; Illas, F.; Wojdel, J. C. Phys. ReV. B 2006, 73, 134202. (13) Zwijnenburg, M. A.; Huenerbein, R.; Bell, R. G.; Cora`, F. J. Solid State Chem. 2006, 179, 3429.

JA0727666

J. AM. CHEM. SOC.

9

VOL. 129, NO. 42, 2007 12589